How to Fix Overexposed Photos (and When You Can't)

An overexposed photo is two different problems wearing the same coat. In one, the picture is simply too bright and everything you need is still in the file. In the other, part of the image hit the sensor's ceiling and was recorded as pure white — and no software, AI included, can retrieve information that was never captured.

Almost every guide to this skips the second half. So: here's how to recover what's recoverable, and how to tell within about ten seconds whether you're wasting your evening.

First, find out what you're dealing with

Open the histogram. Adobe's own Camera Raw documentation puts it plainly: "a histogram with a spike on the right side indicates highlight clipping." A tall wall of pixels jammed against the right edge means clipped highlights, not merely bright ones.

Better still, look at where the clipping is. In Camera Raw and Lightroom, Adobe's docs describe the trick: "While moving the tone controls sliders, hold down Alt (Windows) or Option (Mac OS) to preview where highlights or shadows are clipped. Move the slider until clipping begins, and then reverse the adjustment slightly."

That preview tells you which parts of the frame are actually gone. A clipped sky behind your subject is annoying; a clipped forehead and cheek on a portrait is the shot.

The recovery order that works

Work top to bottom. Each step changes what the next one needs to do.

1. Highlights first, not Exposure. Adobe's own description of the Highlights slider: "Drag to the left to darken highlights and recover 'blown out' highlight details." This targets the bright end without dragging your midtones down with it, which is exactly what dropping Exposure would do.
2. Then Whites. "Adjusts white clipping. Drag to the left to reduce clipping in highlights." Highlights recovers detail; Whites sets where pure white begins. Use the Alt/Option preview here to pull back until the clipping warning just disappears.
3. Exposure only if the whole frame is bright. If the picture is uniformly too light rather than blown in patches, this is the right control. If it's patchy, leave it alone.
4. Blacks and contrast to rebuild. Recovery flattens an image. Deepening the blacks restores the sense of contrast you just removed.
5. Curves last, locally. A gentle S-curve, or a curve applied only to the recovered region through a mask, puts shape back into an area that has gone gray.

The single most common mistake is doing this in the wrong order: dropping Exposure by a stop, discovering the whole photo is muddy, then trying to fix the mud. Recover the top end first.

Why RAW gives you more room

If you shot RAW, you have materially more to work with. Cambridge in Colour's technical reference on the RAW file format explains why: "Most current cameras capture each color with 12-bits of precision (2^12 = 4096 levels) per color channel, providing several times more levels than could be achieved by using an in-camera JPEG" — which is 8-bit, or 256 levels per channel. More recorded levels near the top end means more that survives being pulled back down.

There's a second, subtler advantage. Highlights don't always clip in all three channels at once — a bright red flower may max out the red channel while green and blue still hold data. Adobe's documentation of its legacy Recovery slider describes exactly this: "Camera Raw can reconstruct some details from areas in which one or two color channels are clipped to white." That partial reconstruction is real, and it's why a RAW file often recovers color in an area a JPEG renders as flat white.

The practical consequence: if you're shooting anything high-contrast on purpose, shoot RAW. If you're editing a JPEG someone sent you, expect less.

When it genuinely can't be fixed

Stop and accept the loss when:

• The clipped area is a face. Blown skin has no texture to restore. Filling it means inventing skin, and invented skin on a real person is a different face.
• All three channels are clipped across a large region. No partial-channel reconstruction is available; there is nothing to work from.
• It's a scanned print rather than a digital file. A print that was over-printed or has faded to white in patches never contained the detail in the first place. That's a restoration problem, not an exposure one — our guides to restoring faded photos and restoring old photos cover what's realistic there.

In those cases the honest move is often to crop the blown region out, convert to black and white so a gray patch reads as intentional, or reshoot.

Where AI fits, described accurately

Generative tools can fill a blown-out area with detail that matches its surroundings. That works, and on a sky, a wall, or a background it's often invisible and completely fine.

But it is not recovery. It is generation. The tool is not finding your lost highlight data — it's producing new pixels that look plausible. On a background, nobody is harmed. On a person's face, on a product you're selling, or on a document, the invented detail may be confidently wrong, and you have no way to check it against something that no longer exists.

43frames' photo restoration is built for the restoration end of this: repairing damage, fixing faded color, and enhancing resolution on photos where the problem is age and wear rather than a blown exposure. It costs 2 credits per image, and every new account starts with 10. For a badly clipped digital file, though, honest expectation-setting first: run your exposure recovery in a RAW editor, and use a restoration or upscaling pass afterward to clean up what's left — not instead of it.

If the photo's real problem turns out to be softness rather than brightness, fixing blurry photos is the guide you want, and photo upscaling handles the case where the file is simply too small.

FAQ

Can you fix an overexposed photo? Partly — bright-but-recorded detail comes back with Highlights and Whites. Fully clipped pixels contain no data and cannot be recovered.

Why do recovered highlights look gray? Because there was little tonal variation left in that area. Recover less and add local contrast rather than pushing further.

Overexposed or underexposed — which is easier? Underexposed. Shadows still hold recorded data; clipped highlights don't.

Can AI fix a blown-out photo? It can generate convincing replacement detail, which is invention rather than recovery. Fine on a background, risky on a face.